Cambridge, United Kingdom · Est. 2020

Evidence that holds.

Rigour you can act on.

Data Conscious is a boutique consultancy at the intersection of evaluation, impact, and responsible AI. We help organisations understand what is actually happening — and why.

20

YEARS OF EXPERIENCE

10+

COUNTRY CONTEXTS

3

PRACTICE AREAS

our story

Rigour before reassurance.

Data Conscious was founded on a simple conviction: that the decisions which matter most deserve rigorous evidence, not reassurance.

Working across conflict-affected contexts from DRC to Yemen, Syria to Ukraine, we developed deep expertise in the methods that hold up when conditions are hardest: Outcome Harvesting, Contribution Analysis, Process Tracing, Qualitative Comparative Analysis. These are not just frameworks, they are ways of thinking carefully under pressure.

As AI reshaped how organisations collect and act on data, that rigour found new territory. Our AI governance and assurance practice grew directly from evaluation thinking: not just “does this system work?” but “for whom, under what conditions, and with what risks?”

"Useful evaluation is not about confirming what funders hope. It is about giving decision-makers honest ground to stand on."

 PhD-level research and analytical leadership

Registered with the UK Evaluation Society

Oxford Saïd AI Programme alumni

Based in Cambridge; working globally

How we work: A deliberately small practice. Every engagement our most experience people and produces work we stand behind.

 
 

Neil Dillon

Founder & Principal Consultant

Cambridge, UK

principal consultant

Neil Dillon, PhD

Neil has spent twenty years working at the harder end of evaluation — in conflict and fragility contexts, on sensitive protection mandates, with organisations that needed honest answers more than comfortable ones. His doctoral research sharpened both his theoretical grounding and his scepticism of methods that overpromise.

His field portfolio spans GBV evaluation, IHL promotion, protection monitoring, HDP nexus programming, and large-scale humanitarian response across Sub-Saharan Africa, the Middle East, South and Central Asia, and Eastern Europe. Clients have included UNICEF, WFP, IOM, FAO, FCDO, ICRC, and leading consultancy firms operating in fragile-state contexts.

In recent years Neil has developed a parallel practice in AI governance and assurance — applying the same evaluative logic to the risk assessment and accountability challenges posed by AI systems in high-stakes sectors. He works with organisations navigating the EU AI Act, ISO 42001, and emerging regulatory frameworks, bringing an unusual combination of technical literacy and qualitative depth.

WHO WE WORK WITH

Organisations with

demanding standards.

Our clients tend to share a common profile: they work in complex, high-stakes settings; they have evaluation requirements that exceed what generalist firms can meet; and they value independence.

 

We work directly with UN agencies, INGOs, bilateral donors, and consultancy firms as a specialist sub-contractor. An active expansion brings in development finance institutions, impact investors, family foundations, and corporate clients with developing market and fragile-state exposure.

 

UN AGENCIES & MULTILATERALS

UNICEF

World Food Programme (WFP)

International Organization for Migration (IOM)

Food and Agriculture Organization (FAO)

International Committee of the Red Cross (ICRC)

BILATERAL DONORS

FCDO (UK Foreign, Commonwealth & Development Office)

ECHO/European Commission

SIDA (Swedish International Development Cooperation Agency)

EMERGING CLIENTS

Development Finance Institutions (DFIs)

Impact investors & family foundations

Geopolitical risk consultancies

Corporate clients – fragile-state exposure

Responsible AI / ESG advisory mandates

OUR APPROACH

What working with us looks like.

We are not a firm that scales through volume. Each engagement receives direct principal attention from inception through to final deliverable.

Scoping before designing

Every evaluation starts with a genuine scoping conversation — not a boilerplate inception report. We invest time understanding the decision environment before proposing a design.

Methods selected for the question

We draw from a wide methodological toolkit — quantitative and qualitative, theory-based and participatory — choosing what the question demands rather than defaulting to standard packages.

Specialist collaboration where it adds value

For linguistically or technically complex mandates, we bring in vetted specialists — IHL experts, Arabic translators, subject-matter researchers — always under close quality oversight.

Findings designed for use

Reports, dashboards, and presentations are crafted with their actual audiences in mind. We do not separate analytical quality from communication quality — the two are inseparable.

Proportionate AI integration

We use AI tools selectively and transparently — for translation support, transcript analysis, literature mapping — while maintaining human judgement at every analytical decision point.

GET IN TOUCH

Tell us what you are trying to understand.

We’ll telll you honestly whether we’re the right fit.